An Effective Clustering Algorithm Using Adaptive Neighborhood and Border Peeling Method

Table 1

Performance comparison (datasets flame, R15, compound, and D31).

	Flame		R15		Compound		D31
	ARI	AMI	ARI	AMI	ARI	AMI	ARI	AMI
KNN	0.925	0.773	0.909	0.866	0.904	0.702	0.895	0.897
DBSCAN	0.249	0.298	0.912	0.950	0.424	0.410	0.453	0.782
Border peeling	0.955	0.884	0.993	0.987	0.646	0.765	0.805	0.912
SANBP	0.955	0.857	0.944	0.941	0.946	0.938	0.906	0.941

All the bold values identify the best result.